This survey paper reviews the risk assessment methodology for autonomous vehicles. The paper explains the necessity of real-time risk assessment in conjunction with traditional approaches, especially for highly automated driving. The current methodologies are segmented into different approaches to help identify risk in the aspects of quantitative or qualitative measurements. Further analyses are conducted for each methodology used during development or real-time usage. The outcome of this paper includes a recommended list that addresses each of these methodologies for their suitability toward ISO 26262 and ISO/PAS 21448. In addition, this paper addresses the importance of determinism and uncertainty in different risk assessment methodologies, especially in the usage of AI and machine learning.
Scanning the Issue
IEEE Transactions on Intelligent Transportation Systems ; 23 , 10 ; 16885-16922
2022-10-01
432612 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
IEEE | 2021
|IEEE | 2017
|IEEE | 2017
|IEEE | 2017
|IEEE | 2016
|